Experimental and quantitative analytical studies have shown that human vaginal bacterial communities undergo idiosyncratic changes in species composition and wide fluctuations in the relative abundances of various species. These fluctuations are undeniably associated with specific environmental drivers, however, an understanding of the mechanisms, ecological processes and evolutionary routes behind the genesis of such associations and fluctuations remain an important knowledge gap. The long-term goal of this work is to develop a novel, data driven modeling framework that will allow the identification of fundamental biological processes responsible for the complex microbial community dynamics of the vaginal microbiota and the mechanisms that lead to community shifts. Ultimately, the proposed research would have application in better understanding the causes of conditions such as bacterial vaginosis (characterized by a community shift and dysbiosis), as well as devising strategies to minimize its occurrence and associated risks. To achieve this goal we propose to develop and test a novel stochastic modeling approach to assess the role of environmental drivers in predicting changes in species abundances and composition that shape community-wide stability properties. We will leverage a unique set of over 9,400 samples and associated metadata that were self- collected daily by 135 women for 10 weeks. The rationale for the proposed research is that by embedding stochasticity in community dynamics models a unified framework can be developed that takes into account not only the resilience (return rates to equilibrium) and the reactivity of communities in the face of perturbations, but also the natural and idiosyncratic tendencies to change. Our central hypothesis is that stability properties, diversity and fluctuation regimes of human vaginal microbial communities can be predicted and assessed using a suite of specific stochastic models of the interplay between ecological processes and evolutionary dynamics. We will test this hypothesis by combining time series data of bacterial species composition of human vaginal microbial communities with stochastic models derived from first principles of ecology and evolutionary biology and achieving the following three specific aims: 1) Develop a novel modeling framework that uses a multivariate Ornstein-Uhlenbeck stochastic process to predict the unfolding of complex microbial community dynamics; 2) Characterize the stability properties, architecture, and diversity of ecological interactions within human vaginal bacterial communities; and 3) Probe the ecological and evolutionary mechanisms that lead to community type shifts. The proposed research is innovative and will lead to a deeper understanding of the processes governing species composition, structure and function of bacterial communities in the human vagina. In addition, a theoretical and practical knowledge gap will be filled so that the risk to diseases associated with responses to disturbances of human-associated microbial ecosystems can be better defined.